Importance

• Knowing next season’s expected pitch mixes allows hitters to better prepare over the offseason
  • A hitter coming off a strong season will be scouted more heavily by opposing teams
  • Opposing pitchers will be smarter and throw pitches designed to catch him off-guard
  • By knowing these pitches ahead of time, he can work on what he is more likely to see

The Model (Brief Rundown)

• Uses prior season data to predict pitch mixes faced by a given batter the next year
  • Incorporates both pitch mix faced and performance against different pitch types
  • Accurate within ~6% in total on average (within ~2% on average for each pitch type)
• Model built and tested on the 2021-2023 seasons

The App

• User selects any player with at least 1000 pitches faced and the season to project their pitch mix for (2022-2024 available)
• App outputs: pitch mix and stats for the previous season, projected pitch mix for the upcoming season
• Link to app

Example: Will Benson

Example: Will Benson

Example: Will Benson

Example: Will Benson

• Had an inconsistent season in 2023, sent down to the minors after his first 8 games but returned in May with a vengeance
• Since Benson’s overall season performance was not dramatically spectacular or terrible against any particular pitch type, the model projects he will face a very similar pitch mix in 2024

Example: Spencer Steer

Example: Spencer Steer

Example: Spencer Steer

Example: Spencer Steer

• Had a very successful season in 2023, one of the Reds bright young talents
• Performed very well against fastballs, not as well against breaking balls
• Model projects that Steer will see more breaking balls at the expense of fastballs
  • Important insight for Steer to prepare to see more breaking balls

Conclusion

• Model provides good baseline estimate for pitch mix
  • Not perfect, outside factors could affect pitches seen (changing teams, spot in batting order, etc.)
• Can use app on historical data to test validity